ORACLE RAC修改ora.xxxx.db

修改ora.xxxx.db资源是比较少做的操作，近期因客户修改db_unique_name导致两节点ora.xxxx.db全部显示offline，sqlplus进入发现两实例都是read write状态，此时需要修改ora.xxxx.db资源。

以下是测试信息及过程，修改ora.xxxx.db分为三步骤：

	New	Old
Db_unique_name	orcl	testuni
Instance name	orcl1,orcl2	orcl1,orcl2

• 当前状态

[grid@racnode2 ~]$ crsctl stat res ora.testuni.db -t

--------------------------------------------------------------------------------

NAME           TARGET  STATE        SERVER                   STATE_DETAILS      

--------------------------------------------------------------------------------

Cluster Resources

--------------------------------------------------------------------------------

ora.testuni.db

      1        OFFLINE OFFLINE                             Instance Shutdown  

      2        OFFLINE OFFLINE                             Instance Shutdown

• 删除原来配置

[oracle@racnode1 ~]$ srvctl remove database -d testuni

Remove the database testuni? (y/[n]) y

[oracle@racnode1 ~]$

查看状态

[grid@racnode2 ~]$ crsctl stat res ora.testuni.db -t

CRS-2613: Could not find resource 'ora.testuni.db'.

• 添加新的资源配置

[oracle@racnode1 ~]$ srvctl add database -d orcl -o /u01/app/oracle/product/11.2.0/db_1

[oracle@racnode1 ~]$

查看状态

[grid@racnode2 ~]$ crsctl stat res ora.orcl.db -t

--------------------------------------------------------------------------------

NAME           TARGET  STATE        SERVER                   STATE_DETAILS      

--------------------------------------------------------------------------------

Cluster Resources

--------------------------------------------------------------------------------

ora.orcl.db

1. OFFLINE OFFLINE

[oracle@racnode1 ~]$ srvctl config database -d orcl

Database unique name: orcl

Database name:

Oracle home: /u01/app/oracle/product/11.2.0/db_1

Oracle user: oracle

Spfile:

Domain:

Start options: open

Stop options: immediate

Database role: PRIMARY

Management policy: AUTOMATIC

Server pools: orcl

Database instances:

Disk Groups:

Mount point paths:

Services:

Type: RAC

Database is administrator managed

由于还未添加实例到ora.orcl.db，这里启动实例会收到以下报错

PRKO-3119 : Database orcl cannot be started since it has no configured instances.

• 添加实例到新资源

[oracle@racnode1 ~]$ srvctl add instance -d orcl -i orcl1 -n racnode1

[oracle@racnode1 ~]$ srvctl add instance -d orcl -i orcl2 -n racnode2

[oracle@racnode1 ~]$

查看状态

[grid@racnode2 ~]$ crsctl stat res ora.orcl.db -t

--------------------------------------------------------------------------------

NAME           TARGET  STATE        SERVER                   STATE_DETAILS      

--------------------------------------------------------------------------------

Cluster Resources

--------------------------------------------------------------------------------

ora.orcl.db

      1        ONLINE  ONLINE       racnode1                 Open               

      2        ONLINE  ONLINE       racnode2                 Open

[oracle@racnode1 ~]$ srvctl config database -d orcl

Database unique name: orcl

Database name:

Oracle home: /u01/app/oracle/product/11.2.0/db_1

Oracle user: oracle

Spfile:

Domain:

Start options: open

Stop options: immediate

Database role: PRIMARY

Management policy: AUTOMATIC

Server pools: orcl

Database instances: orcl1,orcl2

Disk Groups:

Mount point paths:

Services:

Type: RAC

Database is administrator managed

状态正常。

总结：

         文章提到的这种操作较少，但是SRVCTL命令很强大，可以再研究下关于其他资源的操作，对于高效率管理集群很有帮助。